Waze navigation app's routing algorithms directed massive volumes of traffic onto narrow residential streets in Los Angeles and other cities, causing safety hazards, traffic congestion, blocked driveways, and multiple accidents in neighborhoods not designed for heavy traffic flow.
Waze, a navigation app owned by Google since 2013, uses AI algorithms to analyze traffic data from its 100 million users worldwide and route drivers through optimal paths to avoid congestion. The app's routing decisions began directing large volumes of cut-through traffic onto narrow residential streets in cities like Los Angeles, where streets designed for 2,000 cars per day were handling over 900 cars per hour, sometimes exceeding capacity by 400 percent. In Sherman Oaks, a yearlong city study found that 86 percent of cut-through traffic was from outside the neighborhood, with over 650 cars per hour on streets designed for local use. This caused multiple safety issues including blocked driveways trapping residents, hindered emergency vehicle access, multiple traffic accidents, road rage incidents, and damaged infrastructure including roads and sewer lines. Los Angeles City Councilman David Ryu sought legal action against Waze in April, citing threats to public safety. Multiple cities implemented traffic restrictions to counter Waze routing, including Leonia, New Jersey closing 60 streets during rush hour, and Brookhaven, Georgia approving road closures. The issue affected numerous cities nationwide including San Mateo, Fremont, and Takoma Park, where one resident was banned from Waze for reporting fake traffic incidents to deter drivers from his street.
Domain classification, causal taxonomy, severity scores, and national security assessments were LLM-classified and may contain errors.
AI systems that fail to perform reliably or effectively under varying conditions, exposing them to errors and failures that can have significant consequences, especially in critical applications or areas that require moral reasoning.
AI system
Due to a decision or action made by an AI system
Unintentional
Due to an unexpected outcome from pursuing a goal
Post-deployment
Occurring after the AI model has been trained and deployed